How to classify Cifar image with Alexnet on Deeplearning4j

Question

I am a beginner to Deeplearning4j, and going to to a testing on Cifar-10 images classify. I just copy the Alexnet from DL4j example(AnimalsClassification.java) like:

    MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
        .seed(seed)
        .weightInit(WeightInit.DISTRIBUTION)
        .dist(new NormalDistribution(0.0, 0.01))
        .activation(Activation.RELU)
        .updater(Updater.NESTEROVS)
        .iterations(iterations)
        .gradientNormalization(GradientNormalization.RenormalizeL2PerLayer) // normalize to prevent vanishing or exploding gradients
        .optimizationAlgo(OptimizationAlgorithm.STOCHASTIC_GRADIENT_DESCENT)
        .learningRate(1e-2)
        .biasLearningRate(1e-2*2)
        .learningRateDecayPolicy(LearningRatePolicy.Step)
        .lrPolicyDecayRate(0.1)
        .lrPolicySteps(100000)
        .regularization(true)
        .l2(5 * 1e-4)
        .momentum(0.9)
        .miniBatch(false)
        .list()
        .layer(0, convInit("cnn1", channels, 96, new int[]{11, 11}, new int[]{4, 4}, new int[]{3, 3}, 0))
        .layer(1, new LocalResponseNormalization.Builder().name("lrn1").build())
        .layer(2, maxPool("maxpool1", new int[]{3,3}))
        .layer(3, conv5x5("cnn2", 256, new int[] {1,1}, new int[] {2,2}, nonZeroBias))
        .layer(4, new LocalResponseNormalization.Builder().name("lrn2").build())
        .layer(5, maxPool("maxpool2", new int[]{3,3}))
        .layer(6,conv3x3("cnn3", 384, 0))
        .layer(7,conv3x3("cnn4", 384, nonZeroBias))
        .layer(8,conv3x3("cnn5", 256, nonZeroBias))
        .layer(9, maxPool("maxpool3", new int[]{3,3}))
        .layer(10, fullyConnected("ffn1", 4096, nonZeroBias, dropOut, new GaussianDistribution(0, 0.005)))
        .layer(11, fullyConnected("ffn2", 4096, nonZeroBias, dropOut, new GaussianDistribution(0, 0.005)))
        .layer(12, new OutputLayer.Builder(LossFunctions.LossFunction.NEGATIVELOGLIKELIHOOD)
            .name("output")
            .nOut(numLabels)
            .activation(Activation.SOFTMAX)
            .build())
        .backprop(true)
        .pretrain(false)
        .setInputType(InputType.convolutional(height, width, channels))
        .build();

When I run the code it threw an exception say there are some problems with "layer-9" configuration on new int[]{3,3}, it should be greater than 0 and less than pHeight + 2*padH. When change the weight*height from 32 * 32 to 100*100 in java code, it ran properly, but I and not should the result is good. So I am a little bit confused on the layer configuration on alexnet deal with 32*32 images.

Answer 1

That isn't going to be the right example to use. Please wait till we finish out our new model import from keras instead. That will also include the pretrained models.